[mythtv-users] Does higher bitrate and resolution benefit HDTV?

Steven Adeff adeffs.mythtv at gmail.com
Fri Dec 15 17:53:25 UTC 2006


On 12/14/06, Gary Taylor <squeaky at sdf.lonestar.org> wrote:
> I want to use Myth to archive my analog 8 mm camcorder home movies to DVD.
> Right now I only have a 20 year old TV, but eventually I imagine it's
> inevitable that I'll get an HDTV.  Will saving my home movies using the higher
> resolution and bitrate that the HQ profile in Archive Files provides, provide
> me any advantage when viewed on HDTV when I get one?
>
> If saving my movies in the higher resolution and bitrate will not do me any
> good then I'll save them in 352x480 and there is no need to read further.
> Otherwise....
>
> I've been experimenting with tapes that contain about 2 hours 4 minutes worth
> of material and with some adjustments to mythburn.py(dvdrsize=(4482,8152))
> and/or ffmpeg_dvd_ntsc.xml(bitrate 8000), I've been able to make them fit onto a
> DL-DVD and they play in my DVD player. If I don't make these changes the
> resulting .iso image is too big and K3b complains.
>
> It seems that hacking mythburn.py and ffmpeg_dvd_ntsc is undoing someones well
> thought out plans.  Short of cutting my home movies shorter, is there a better
> way I should be doing this?  I haven't really looked at how to edit the raw
> files from my PVR-350 so maybe it's the best way to go. I seem to recall a
> thread about the audio and video getting out of sync when editing like that,
> but am unsure.
>
> I like using mythburn.py from the command line because it allows me to automate
> everything starting from the raw capture to the DVD burn, but again I feel like
> I'm bypassing someones well thought out GUI design.
>
> I'm capturing from a PVR-350 using 720x480
> Debian Etch
> Using MythTV 0.20 from the debian-multimedia repository (Thanks Christian Marillat)

The simple answer is, yes, always. For what your doing, I would
suggest capturing at 720x480 with the maximum bitrate possible
(normally over kill yes...), then using something like avidemux, or
similar windows software, doing any editing you wish to do, both to
the time line and correction work(color, etc), then re-encoding the
video from there.
As to how to choose your final resolution, this can be tricky, and
more time consuming than just leaving it at 720x480. You have to test
and see what the effective resolution of the capture is. It very well
may be that saving it at 1/2 resolution is more than enough for the
quality of your film. The basic way to figure this out is to lower the
resolution in the encoding process, then on playback resize to the
original size and do a visual comparison. One issue you will find is
that at these lower resolutions this doesn't work all that wel as the
resize will always come out different. What you  have to pay more
attention to than the exact pixel space is whether the difference in
quality warrants the space saved by the lower resolution. From there,
bitrate is a similar story. Keep going lower until your happy with the
outcome. This is where the idea of a encoding to a "quality" level
comes in. The idea being the software will determine the bitrate of
the output based on your quality level. MPEG's rough version of this
"quality" settings is the quantization value. DVD's and HD-DVD's are
encoded at a Q=0(or 0,1), the highest quant level possible(0,1 having
to do with how the quant level for different types of frames are
chosen).

The more complex answer on what to capture at has to to do with what
you plan on exactly doing later. All forms of analog video capture
(film) have what is termed an "effective" resolution. that is to say
that there is some, in this case chemical, structure that makes up the
smallest piece of information that can be stored. 8mm and similar
analog digital sources are limited by either the CCD or the analog
algorithm used. So it would make sense to use the highest resolution
capable by the source as the capture. This is what is currently done
when converting movie film to digital, they use the maximum effective
resolution of the source to perform the capture. Anything more is not
only a waste, but can introduce sampling error which will cause the
captured video to not accurately represent the source. Films are
usually captured at ~4x the "effective" resolution for editing, after
the editing, postprocessing, CGI, etc are performed, the film is then
resized to the proper resolution for distribution. Doing this gives
the engineers more control over the quality of the final output of
their editing process. Films being captured for mass media release
(DVD, HD-DVD, archival) are done at their "effective resolution".

-- 
Steve
Before you ask, read the FAQ!
http://www.mythtv.org/wiki/index.php/Frequently_Asked_Questions
then search the Wiki, and this list,
http://www.gossamer-threads.com/lists/mythtv/
Mailinglist etiquette -
http://www.mythtv.org/wiki/index.php/Mailing_List_etiquette


More information about the mythtv-users mailing list